1. Field of the Invention
The present invention relates to a solid state image sensor and a method for fabricating the same.
2. Description of Related Art
A CCD (charge coupled device) type solid state image sensor is so constructed that light is incident on an n-type semiconductor region formed in a surface of a P-type silicon substrate and an image signal is obtained from a signal charge in the -type semiconductor region.
Referring to FIG. 17, there is shown a sectional view of one example of the prior art solid state image sensor, which includes tranfer electrodes 3 formed on a P-type silicon substrate 11 with a sixth insulating film 12f formed of a silicon oxide film being interposed between the tranfer electrodes 3 and the P-type silicon substrate 11. In a surface of the substrate 11 between each pair of transfer electrodes 3, an n-type semiconductor region 17 is formed to constitute a photoelectric conversion region. Above this n-type semiconductor region 17, an aperture is formed in a light block film 16 formed of aluminum or tungsten. A passivation film 18 is formed to cover the light block film 16. Furthermore, on the surface of the substrate 11 under the transfer electrode 3, a second n-type semiconductor region 21 is formed to constitute a transfer region, and one end of the second n-type semiconductor region 21 is separated from the n-type semiconductor region 17. A p.sup.+ semiconductor region 26 is formed between the other end of the second n-type semiconductor region 21 and the n-type semiconductor region 17 in order to isolate pixels from one another. In the following, this prior art will be called a first prior art.
In the structure of the first prior art shown in FIG. 17, however, the loss of an incident light is large because of reflection at the surface of the P-type silicon substrate 11, and therefore, a satisfactory sensitivity cannot be obtained.
In order to overcome this problem, for example, Japanese Patent Application Pre-examination Publication No. JP-A-04-206571 (an English abstract of which is available and the content of the English abstract is incorporated by reference in its entirety into this application) proposes to form an antireflection film in the photoelectric conversion region. In the following, the prior art typified by JP-A-04-206571 will be called a second prior art.
Now, the second prior art will be described with reference to FIG. 18. In FIG. 18, elements corresponding to those shown in FIG. 17 are given the same reference numbers.
In this second prior art, for example, an n-type semiconductor region 17 becoming a photoelectric conversion region for obtaining a signal charge, and a second n-type semiconductor region 21 becoming a transfer region for transferring the signal charge read out from the n-type semiconductor region 17, are formed in a principal surface of the P-type silicon substrate 11. The n-type semiconductor region 17 and the second n-type semiconductor region 18 are formed by for example an impurity diffusion. Incidentally, pixels are isolated from one another by a p.sup.+ semiconductor region 26.
Furthermore, a seventh insulating film 12g formed of a silicon oxide film is formed on the P-type silicon substrate 11. On the silicon oxide film 12g, there is formed an antireflection film 15 formed of a silicon nitride film having a refractive index larger than that of silicon oxide but smaller than that of silicon. The refractive index of the silicon oxide is about 1.45, and the refractive index of the silicon nitride is about 2.0. Film thicknesses of the seventh insulating film 12g and the antireflection film 15 are not greater than 600 .ANG., respectively, and preferably on the order of 250 .ANG. to 350 .ANG., respectively.
By setting the film thicknesses, an antireflection film having a relatively flat spectral characteristics in a visible light region can be obtained. Thus, by setting the film thicknesses of the seventh insulating film 12g and the antireflection film 15 to an appropriate thickness, the reflection factor is suppressed to 12% to 13% at average. Since the incident light was reflected about 40% in the prior art P-type silicon substrate, the reflection factor can be reduced to about one third.
A polysilicon layer functioning as a transfer electrode 3 is formed through the sixth insulating film 12f on the silicon oxide film 12g and the antireflection film 15 above the transfer region. The transfer electrode 3 is coated with an eighth insulating film 12h formed of a silicon oxide film, and furthermore, is coated with the light block film 16 in order to block the incident light. The light block film 16 is formed of for example aluminum. An aperture is formed on the light block film 16 positioned above the n-type semiconductor region 17 so that the light block film 16 faces onto the n-type semiconductor region 17 in the aperture. The light block film 16 is overcoated with a passivation film 18. With this arrangement, a high sensitivity can be realized.
However, the above mentioned structure has the following problems:
A method for effectively reducing a dark current in the solid state image sensor is to diffuse hydrogen, as disclosed in for example Japanese Patent Application Pre-examination Publication No. JP-A-06-209100 (an English abstract of which is available and the content of the English abstract is incorporated by reference in its entirety into this application).
In the structure in accordance with the second prior art, it is not possible to sufficiently perform the terminating of dangling bonds at a silicon interface by hydrogen in a final sintering step, which is effective in reducing the dark current. The reason for this is as follows: When the sintering is executed after the antireflection film of the silicon nitride film is formed, hydrogen is prevented from reaching the silicon interface by action of the silicon nitride film of the antireflection film.
Incidentally, the solid state image sensor disclosed in JP-A-06-209100 has no antireflection film, and JP-A-06-209100 does not disclose a method for reducing the dark current when the antireflection film is provided.
A second problem is that since a driving characteristics of the transfer region is limited, it becomes difficult to lower a driving voltage of the transfer electrode. In order to increase the sensitivity in the visible light region, it is necessary to form the antireflection film having the film thickness on the order of 300 .ANG. to 500 .ANG.. If the antireflection film of this film thickness is actually formed on the whole surface, the film thickness is the same between the photoelectric conversion region and the transfer region, and therefore, the film thickness in the tranfer region is also on the order of 300 .ANG. to 500 .ANG.. On the other hand, in order to drive the transfer electrode with a low voltage, it is necessary to make the capacitance directly under the transfer electrode as small as possible. For this purpose, it is necessary to make the oxide film directly under the transfer electrode as thick as possible. Because of this, it is difficult to lower the driving voltage of the transfer electrode.